T-06

Sensory evaluation

Trained-panel testing for the questions no instrument can answer: whether anybody can actually tell the difference.

What it is: Difference, descriptive and acceptance panels
Who it is for: Anyone reformulating, changing supplier or fighting a complaint
Turnaround: 3–10 days

WHAT THIS IS

What this is, and why it matters

Sensory evaluation uses people as the measuring instrument, under controlled conditions and to published ISO methods. Difference testing — triangle, duo-trio, paired comparison — establishes whether a change is detectable at all. Descriptive analysis puts a trained panel on defined scales to describe how two products differ, attribute by attribute. Acceptance testing puts the product in front of consumers to establish whether they like it. Taint and off-note screening investigates a specific complaint.

The discipline exists because the most commercially important questions in food have no instrumental answer. Whether a cheaper ingredient is noticeable, whether a new supplier has changed the product, whether a batch complaint is a real defect or one unhappy person, whether a product at the end of its stated life is still acceptable to eat — none of these are settled by a chromatogram. What makes a sensory result evidence rather than opinion is structure: a defined method, a stated panel size, a scoring scale, and a statistical result rather than a show of hands.

What this lane covers

Difference testing, descriptive analysis, acceptance panels
Reformulation, supplier change, benchmarking, complaints

TRIGGERS

When you need it

You are changing something

A reformulation, a cost reduction, a packaging change or a new supplier all raise the same question: will the customer notice? A difference test answers it before the change ships rather than after the complaints arrive.

A complaint needs adjudicating

A taint or off-note complaint is either a real defect or an outlier, and the difference decides whether you have a batch problem. A screening panel establishes which, and gives you something defensible to put in the response.

A shelf-life study needs its sensory leg

Products usually become unacceptable to eat well before they become unsafe to eat. A shelf-life study without a sensory checkpoint has dated the pack on the wrong criterion.

CATALOGUE

Sensory — parameters

Parameter	Matrix	Reference method	Turnaround
Triangle test (difference)	Any food or beverage	ISO 4120	3–5 days
Duo-trio test	Any food or beverage	ISO 10399	3–5 days
Paired comparison	Any food or beverage	ISO 5495	3–5 days
Quantitative descriptive analysis	Reformulation, benchmarking	ISO 13299	7–10 days
Hedonic acceptance panel	Consumer-facing product	ISO 11136	7–10 days
Taint and off-note screening	Complaint investigation	ISO 13302	3–5 days
Shelf-life sensory tracking	Product over storage	ISO 16779	Per study plan

Method references and turnaround times shown are indicative and are confirmed in writing before any sample is accepted.

Figure · Published turnaround spread

Published turnaround spread, grouped from the catalogue above — 6 of 7 lines carry a working-day figure. The other line runs to a study plan, an interval or the visit itself, and is not drawn.

PROCESS

How it works

Send the brief before the sample
The question you actually need answered, in your own words — can they tell, how does it differ, do they like it, what is this off-note. That question selects the method, and choosing the method first is how sensory work ends up answering something you did not ask.
Sample requirements confirmed in writing
Sample count is higher than in other disciplines because each panellist needs their own portion, and coded samples usually need to be prepared identically. Quantities, coding and any preparation instruction are confirmed before you send anything.
On the bench, to a published method
Every determination runs to the ISO, AOAC or APHA method named in the catalogue above, and that reference travels onto the report line rather than sitting in a footnote. Turnaround runs from acceptance at the laboratory, not from the day the sample left you.
Reported so the result can be acted on
Results are reported with the method, the panel size and the statistical outcome, including whether the result reached significance. A difference test that finds nothing is a real answer, and it is reported as one rather than dressed up.

What you get back

A report stating the method, the panel size, the scoring scale, the result and its statistical significance — so it can be attached to a change control or a complaint file and still mean something to a reader who was not in the room.

Service facts

Parameter	Value	Note
Live service status	[VERIFY: Which testing services are actually LIVE]	Published as orderable only when the lab confirms it.
Accreditation position	Accredited. Certificate N-T-00607 covers food and agricultural microbiological testing to ISO/IEC 17025:2017.	Stated plainly, never implied by a badge.
Sample requirements	Confirmed to you in writing before dispatch. Container, quantity, temperature and hold time change with the parameter and the matrix, so they are issued per request rather than published as one list.	Per-parameter detail publishes with the confirmed method list.

Q+A

Common questions

Sensory

What is the difference between a difference test and an acceptance test?

A difference test answers one narrow question — can people tell sample A from sample B. An acceptance test asks whether people like the product. Descriptive work sits between them: it says what the difference actually is, in trained, repeatable terms.

I changed an ingredient supplier — which test do I need?

Start with a difference test against your current product. If no difference is detected, the change can proceed on evidence. If a difference is found, descriptive work tells you what changed and by how much — which is what you take back to the supplier.

Can my own staff be the panel?

Not for formal work. Staff know the product, know what changed and often know what result is wanted — all three bias the outcome. A trained panel is screened and calibrated so the result stands up when a customer questions it.

Will an acceptance test tell me the product will sell?

No — it tells you whether the people who tasted it liked it, under controlled conditions. That is a real and useful signal, but it is not a market forecast, and the report will not pretend otherwise.

Isn’t sensory testing just opinion?

The methods exist precisely to control opinion — defined procedures, screened panellists, blinded samples and statistics on the outcome. The report states the method and the confidence in the result, the same way an instrument test states its method.

Send the sample question or the instrument list. It is read by a bench professional, not routed through a ticket queue. Confirmed phone, mailbox and address publish here at launch.

Submit a sample Contact

CONTINUE